I work in marketing for a major corp. My teams manage 100+ ecommerce sites in 15 languages; we send out millions of marketing/promo emails each month; we manage product-search taxonomies, geographic hierarchies, keywords, etc. Even though large databases are our workhorses, we also transform tens of thousands of text content files (XML primarily) and validate data across plumbing junctions-- ETL jobs, SOA, web APIs and so-forth.

My teams are business users-- web-savvy marketers, designers and data analysts -- i.e., a semi-technical group. Some can write javascript and/or SQL, but they aren't programmers. We need very simple, short, readable scripts. Most languages require regular expressions, which are not a good fit for my teams unless the task is fairly basic. Rebol's Parse DSL is probably the best regular expression replacement I've found in my career. We basically have no need for RE's ever.

Rebol (and Red) have plenty of other uses -- and we've built some nice desktop GUI apps. But our main needs are around data-munging/wrangling -- scanning tens of thousands of files and/or webAPIs and parsing/extracting/transforming that data to make our many haystacks needle-free. For us, Rebol/Red have the simplest tools that work well for this job.

I imagine in the future, with greater emphasis on knowledge worker roles (such as data analysts), we may see the emergence of even simpler languages than available today-- My $.02 -- I just don't think R/Python/Ruby/Perl are quite simple enough for semi-technical staff and non-programmers.